The Diabetes Prevention Program (DPP) randomized 3,234 subjects at high risk of type 2 diabetes (T2D) to one of three arms: placebo, metformin 850 mg twice daily or an intensive lifestyle intervention (a fourth arm of 585 subjects assigned to troglitazone treatment was stopped prematurely because of one incident of fatal hepatotoxicity). Over a 4-year period, close to 40% of subjects in the placebo group developed T2D; there was a significant reduction in the incidence of T2D among both the metformin and the lifestyle intervention groups. We propose to explore whether common genetic variation at a number of candidate genes might help predict progression to T2D or response to any of the above interventions. We have selected up to 66 genes which have been implicated in glucose homeostasis. These include genes that cause monogenic forms of T2D or obesity, genes that encode targets of hypoglycemic medications, genes involved in cellular energy regulation (including the mitochondrial genome), genes that participate in metabolic endocrine networks and genes affected by the human response to exercise. The haplotype structure of all of the above genes will be fully characterized with the aid of the data provided by the HapMap Project, complemented by our own genotyping and sequencing efforts. Haplotype tag single nucleotide polymorphisms will be genotyped in the DPP cohort. The impact of a specific allelic variant on the development of T2D will be assessed in each of the intervention arms. Secondary endpoints will include measures of insulin secretion and insulin resistance based on the initial and final oral glucose tolerance test. If successful, the proposed research should help elucidate the genetic heterogeneity of T2D and obesity, as well as lay the foundation for pharmacogenetic studies of treatment response and prevention in T2D.